abstract class %iKnow.Source.Processor extends %Library.RegisteredObject
This is the base class for any Processor implementations. Processor classes are developed to read source data as referenced (or stored temporarily) by %iKnow.Source.Lister classes in their ExpandList() methods. It is up to the Lister to make sure the references they store in this stage refer to the proper Processor implementation and contain the details it needs to find and read that source while executing FetchSource().
Each Processor implementation tackles a specific type of location where source data can be read from, such as a filesystem for %iKnow.Source.File.Processor or a global for %iKnow.Source.Global.Processor.
Processor objects are handled by the loading framework automatically and custom code should never instantiate a Processor object or call any of its methods directly.
This method is called from ProcessListElement() and ProcessVirtualListElement() right before calling on the Indexer to process the data read by this Processor instance. It returns the language (two-letter ISO code) for the source being processed, if there is any means to derive it from known facts about the source. If no language can be derived with enough certainty, this method returns "".
The default implementation looks if there is a $$$IKPLANGUAGEFIELDNAME parameter defined pointing to a metadata field containing the language code. Subclasses may override and/or extend this method if they have access to additional information.
Subclasses should override this method to identify those metadata keys that are calculated or extracted by this Processor implementation and will be registered for every source processed by calling SetCurrentMetadataValues() during the FetchSource() execution.
This is the primary method that will be called from the %iKnow.Source.Loader for each source identified by the %iKnow.Source.Lister objects involved in this loading process. It will buffer what's read in the FetchSource() implementation, trigger the %iKnow.Source.Converter if one is configured, and then instruct the Indexer to process the resulting plain text, directing its output to the appropriate DirectInput object.
Also, this method will gather any metadata set by the Lister, combine it with metadata calculated by the Processor or Converter and push it to the DirectInput object as well.
Registers the Converter to be used by this Processor onwards.